Extracting Structural Motifs from Pair Distribution Function Data of Nanostructures using Explainable Machine Learning

Andy Sode Anker, Emil Thyge Skaaning Kjær, Mikkel Juelsholt, Troels Lindahl Christiansen, Susanne Linn Skjærvø, Mads Ry Vogel Jørgensen, Innokenty Kantor, Daniel Risskov Sørensen, Simon J. L. Billinge, Raghavendra Selvan, Kirsten Marie Ørnsbjerg Jensen

Keywords: Explainable Machine Learning, X-ray scattering, Pair Distribution Function, Nanoparticle characterization, Materials Chemistry

TL;DR: We present our newly developed algorithm, Machine Learning based Motif Extractor (ML-MotEx), and use it on 4 examples of experimental pair distribution function datasets to assign each atom an importance value for the fit quality.

Abstract: Characterization of material structure with X-ray or neutron scattering using e.g. Pair Distribution Function (PDF) analysis most often rely on refining a structure model against an experimental dataset. However, identifying a suitable model is often a bottleneck. Recently, new automated approaches have made it possible to test thousands of models for each dataset, but these methods are computationally expensive, and analysing the output, i.e., extracting structural information from the resulting fits in a meaningful way is challenging. Our Machine Learning based Motif Extractor (ML-MotEx) trains an ML algorithm on thousands of fits, and uses SHAP (SHapley Additive exPlanation) values to identify which model features are important for the fit quality. We use the method for 4 different chemical systems including disordered nanomaterials and clusters. ML-MotEx opens for a new type of modelling where each feature in a model is assigned an importance value for the fit quality based on explainable ML.

Paper Track: Papers

Submission Category: Automated Material Characterization